[PyMOL] (almost) using APBS within PyMOL

2004-09-07 Thread Michael Bovee

Okay, I think I'm getting real close to having this working
(Mac G4 AlBook, OSX 10.3.5, Apple X11, X11PyMOL 0.97)

Here's my simple-minded approach so far, to get electrostatics in PyMOL  
instead of the Honig Lab's online GRASS setup (which is also good).


1) maloc and apbs have been successfully compiled from source and  
installed in /usr/local


2) M. Lerner's APBS Tools plugin 'About' tab says to load a structure;  
the 'Visualization' tab says to load a molecule and a map  So, I just  
loaded a pdb file like I usually do, and displayed as cartoon


3) set the correct location of the apbs binary (/usr/local/bin/apbs),  
clicked 'Set grid', then clicked 'run APBS'


4) those pink anchor handles in pymol started flashing around my  
protein cartoon, then a pause, then the cartoon disappeared and the  
pink anchors flashed around some more, then all went black. Nothing  
happens when I click 'Update' under the Visualization tab.


5) PyMOL output:
 
_

coarsedim is [155.64339675903321, 159.7681230545044, 272.30599403381348]
finedim is [111.55493927001953, 113.98124885559082, 180.17999649047852]
center is [128.19829940795898, 61.232932090759277, 21.427999496459961]
finegridpoints is [225, 257, 385]
radiobutton said to generate it Use PyMOL generated PQR and PyMOL  
generated Hydrogens and termini so i am returning pymol-generated.pqr
radiobutton said to generate it Use PyMOL generated PQR and PyMOL  
generated Hydrogens and termini so i am returning pymol-generated.pqr

 WARNING: 48 atoms did not have formal charges assigned
 WARNING: 96 atoms did not have properties assigned
ObjectMapLoadDXFile-Error: Unable to open file!
 
_


supposing the WARNINGs are harmless, I'm not sure what to try next.  
Maybe I need the electron density map after all?


--Michael




Re: [PyMOL] installing APBS on PBG4

2004-09-02 Thread Michael Bovee

On Sep 2, 2004, at 2:03 PM, Alex W. Smith wrote:

Michael:
I followed the directions for installing from source and used Bill 
Scott's tweak (both links below).  Everything seems to be running 
fine.  I am a bit short of memory using my Al G4PB, but everything 
configured and complied with no problems.  I have only used APBS with 
the Pymol plugin but it seems to run fine.  Slow but fine.


Hi Alex, and anyone else who might have a suggestion here;
I am also trying to use APBS for electrostatics in PyMOL. Running a G4 
AlBook, 10.3.5, Apple's X11, latest fink, XCodeTools 1.0 (I havent 
upgraded to 1.5 yet).  But I got the personal attention of my Unix guru 
for a while today and made some progress installing from source. He 
helped me get maloc installed in /usr/local by using

./configure --prefix=/usr/local


Then I started on APBS, by following Bill Scott's tips, summarized here:
1. start up zsh (to make use of its recursive globbing capability)
2. ./configure
3a. perl -pi -e 's|\-L/usr/local/|\-lcc_dynamic \-L/usr/local/|g' 
**/Makefile

3b. make
3c. make install


I got the following errors that stopped the make:

g77 -g -O2 -g -O2 -o apbs main.o routines.o  
-L/Users/mbovee/lib/powerpc-apple-darwin7.5.0 
../src/aaa_lib/.libs/libapbs.a -L/Users/mbovee/Desktop/apbs-0.3.1/lib 
/usr/local/lib/powerpc-apple-darwin7.5.0/libmaloc.a 
/usr/local/lib/powerpc-apple-darwin7.5.0/libblas.a 
/usr/local/lib/powerpc-apple-darwin7.5.0/libvf2c.a -lm
/usr/bin/ld: warning -L: directory name 
(/Users/mbovee/lib/powerpc-apple-darwin7.5.0) does not exist
/usr/bin/ld: warning -L: directory name 
(/Users/mbovee/Desktop/apbs-0.3.1/lib) does not exist

/usr/bin/ld: Undefined symbols:
restFP
saveFP
collect2: ld returned 1 exit status
make[1]: *** [apbs] Error 1
make: *** [all-recursive] Error 1


doing a little more checking on these undefined symbols, I find:

$ nm *.a |grep restFP
 U restFP
 U restFP
$ nm *.a |grep saveFP
 U saveFP
 U saveFP

Can anyone tell me where these symbols are defined, or what I screwed 
up? Maybe I should just start over and let the scripts install 
everything in my homedir automatically...


--Michael

[PyMOL] color individual residues on a surface?

2004-07-23 Thread Michael Bovee

Hi PyMOL'ers

I'm trying to figure out how to do something that looks like the 
following figure:

http://www.uvm.edu/~mbovee/structures/surface.png

It seems any individual residues I color always map onto the cartoon 
ribbon representation rather than onto the surface representation.  Is 
this currently possible in PyMOL to do what is pictured in the link 
above?


Thanks,
--Michael Bovee, Ph.D.
Dept. of Biochemistry
University of Vermont




[PyMOL] problem with dist

2004-05-17 Thread Michael Bovee
Hi,I'm running MacPyMOL 0.95 on a G3 Lombard/BronzeKeyboard 
Powerbook, OSX 10.2.8.


The ref manual says to measure distances you can ctrl-shift left, and 
then right click to select the positions to measure. I can ctrl-shift 
left click to select the first atom, but ctrl-shift right click doesnt 
do anything. My workaround is to ctrl-shift left click twice, which 
displays the path to the atom in the command history window. I can then 
copy paste them in behind a dist command.


Can anyone else confirm this behaviour?

Thanks,
--Michael




Re: [PyMOL] question on differential coloring of protein chains

2004-02-18 Thread Michael Bovee

On Tuesday, February 17, 2004, at 02:57  PM, ALEX DAJKOVIC wrote:


hello-
i am new to pymol and am trying to figure out how to differentially
color different chains (i.e. different proteins) in the structure i am
viewing.  the structure i am working with is actually a structure of 
two

proteins that were co-crystalized and i would like assign different
colors to them.
thanks for your help.
alex dajkovic


Hello Alex, I think what you want is to use the 'select' function in 
combination with the 'resi' function. This is documented in the 
reference manual.
Following is a helpful summary that was sent to me by Robert L. 
Campbell, Ph.D.


Let's say I wanted to create a single residue selection for Lysine 465 
(and I choose to call the selection 'K465'), so that I could color and 
display it the way I wanted: (insert your particular atom coordinates 
file name where you see square brackets below)

___
select K465, ([filename] and chain A and resi 465)

or you could shorten the selection string:

select ak465, (abc  c. a and r. 465)

or use the shorthand notation:

select ak465, /abc//a/465
___
˜
Now to create multiple residue selections, just separate the start and 
end residue numbers by a colon, I think, following the word 'resi'


Once you create your unique selections they show up in the right panel 
list and you can further manipulate their display properties as you see 
fit.


Regards,
--Michael




[PyMOL] cannot print most of user manual pdf

2003-12-16 Thread Michael Bovee

Hi there,
Trying to make myself a desk copy of the user manual is proving to be a 
pain.
I can print the text of the table of contents, but anything with screen 
shot graphics never prints from our new HP2300 (48MB of memory). The 
refman pdf printed fine, but this userman pdf file seems corrupted in 
some way. It is only 1.4 MB, but I cannot even print a selection of 4 
pages onto a one sheet page layout.


Is this a known problem, or just MY problem?  :0)
Thanks,
--Michael




[PyMOL] assign secondary structure?

2003-12-11 Thread Michael Bovee

Hi,
I have a 'raw' pdb file of atom coordinates only, and I'm hoping to use 
the dss command in pymol to get helix and strand designations that I 
can paste back into the header space of the pdb so that I can model 
fancy helices and strands and so forth to get a nice picture in pymol. 
Then I can superpose features with homologous proteins for which this 
header info is available. I just didnt want to modify or create this 
header info 'by hand' if I didn't have to...


But, I can't get dss to work in pymol (even the refman warns that this 
algorithm has not been rigorously validated). The refman says USAGE is 
dss selection, state
Does this mean I need to first make a selection that encompasses some 
part (or whole) of the molecule, or should this work on the entire pdb? 
 I don't know what 'states' are so I just opted for '0' which is 
supposed to mean 'all states'.


Thanks,
--Michael Bovee




[PyMOL] cannot 'fit' a particular pdb file

2003-12-02 Thread Michael Bovee

Hi,
I have loaded 3 pdb files of homologous forms of an enzyme. The 'fit' 
command works for superimposing/aligning two of them, but the third 
file always gives a No Atoms Selected error. Is there something in the 
pdb file I can fix? As a general rule now, I Hide Everything at first 
and then Show Lines as I begin with a new PyMOL session. I can see all 
three molecules, just can't align the third one (EcHRSade) with the 
first two (TtHisRS and SaHRS).


PyMOLfit TtHisRS, SaHRS
 Executive: RMS =4.222 (718 to 718 atoms)

PyMOLfit TtHisRS, EcHRSade
ExecutiveRMS-Error: No atoms selected.
PyMOLfit EcHRSade, TtHisRS
ExecutiveRMS-Error: No atoms selected.
PyMOLfit EcHRSade, SaHRS
ExecutiveRMS-Error: No atoms selected.

Thanks,
--Michael




[PyMOL] aa residues disconnected from 'fancy helices'?

2003-11-07 Thread Michael Bovee

Hi all,
I'm just starting to try out PyMOL (fink package 0.90-2, running 
through Apple's X11beta2, on a Lombard G3PB, MacOS 10.2.4).
Its much easier to learn that I thought it would be! Well done Warren, 
and codevelopers!  One puzzling thing I've noticed: when I build a 
model of my protein of interest using 'set cartoon_fancy_helices,1' and 
then show sidechains for several key catalytic residues, I find that 
the amino acids that reside along beta sheet or alpha helix regions 
appear properly connected to the alpha carbon backbone representation, 
but sidechains of amino acids that reside on 'strand' regions  appear 
to float in midair and do not connect to the strand backbone. Anyone 
else seen this behaviour?


Thanks!
--Michael Bovee
University of Vermont




Re: [PyMOL] aa residues disconnected from 'fancy helices'?

2003-11-07 Thread Michael Bovee

Friday, November 7, 2003, 8:50:13 AM, you wrote:
MB ...I built a model of my protein of interest using 'set 
cartoon_fancy_helices,1'...but sidechains of MBamino acids that 
reside on 'strand' regions appear to float in midair and do not 
connect to the strand MBbackbone...


On Friday, November 7, 2003, at 12:37  PM, Jacob Corn wrote:

snip  turn off Smooth Loops under the Cartoon menu...
Jacob


Ayup, that did the trick. Thanks!
--Michael